Two-dimensional (2D) hexagonal boron nitride materials are isomorphs of carbon nanomaterials and hold promise for electronics applications owing to their unique properties. Despite the recent advances in synthesis, the current production capacity for boron nitride (BN) nanostructures is far behind that for carbon-based nanostructures. Understanding the growth mechanism of BN nanostructures through modeling and experiments is key to improving this situation. In the current work, we present the development of a ReaxFF-based force field capable of modeling the gas-phase chemistry important for the chemical vapor deposition (CVD) synthesis process. This force field is parameterized to model the boron nitride nanostructure (BNNS) formation in the gas phase using BN and HBNH as precursors. Our ReaxFF simulations show that BN is the best of these two precursors in terms of quality and the size of BNNSs. The BN precursors lead to the formation of closed BNNSs. However, BNNSs are replaced with complex polymeric structures at temperatures of 2500 K and higher due to entropic effects. Compared to the BN precursors, the HBNH precursors form relatively small, flat, and low-quality BNNSs, but this structure is less affected by temperature. Additives like H2 significantly affect the BNNS formation by preventing closed BNNS formation. Our results show the ReaxFF capability in predicting the BN gas-phase chemistry and BNNS formation, thus providing key insights for experimental synthesis.